"Ozempic Butt": What Actually Happens, and What the Evidence Says You Can Do

The short version

"Ozempic butt" is a nickname for something real: after fast, large weight loss the buttocks and outer hips lose volume, flatten, and often sag, and the change reads as disproportionate because it lands in a depot women carry more of to begin with.

Two things are worth separating immediately, because almost nothing written about this separates them.

It is not a drug effect on your buttocks. It is what happens to a fat depot when a lot of fat leaves quickly. The best data available show the proportion of weight lost as fat versus lean tissue is roughly the same on the drug as on placebo — what the medication changes is how much comes off, and how fast¹.

Nobody has measured it. Not once. There is no trial reporting buttock volume, gluteal contour, or hip circumference as an outcome in people taking these medicines. Every figure you will see attached to this phrase has been inferred from whole-body scans. That absence is the single most important fact on this page, and I will come back to it.

What the body-composition data actually show

The cleanest source is the DXA substudy of SURMOUNT-1, in which 160 of the trial's participants — 73% of them women, mean weight 102.5 kg — were scanned at baseline and again at week 72¹.

Look at the last row rather than the first. Of the weight lost, roughly three quarters was fat and one quarter lean tissue — and that ratio was the same in the placebo group¹. The people losing 5% of their body weight without the drug were losing it in the same proportions as the people losing 21% with it.

A 2026 systematic review in the Annals of Internal Medicine pooled 35 randomized trials of liraglutide, semaglutide, tirzepatide and dulaglutide and reached the same place from a wider base. Across incretin groups, the median proportion of total weight loss attributable to muscle-related indices was 28.3% (IQR 15.9% to 39.9%), exceeding the review's prespecified benchmark in about two thirds of studies — but the benchmark was also exceeded in nearly half of the non-drug interventions that produced weight loss². The authors' own limitation is worth carrying: body-composition methods were too heterogeneous to meta-analyze, and no study reported objective physical function².

So the honest framing is not "the drug eats your muscle and your backside." It is: fast weight loss costs lean tissue and fat together, in a fairly fixed ratio, and these medicines produce far more of it — around 15% average weight loss with semaglutide in its pivotal trial³ — than diet and exercise usually do.

Why the buttocks specifically

Because of where women store fat, and what that fat is for.

Body fat distribution is strongly sex-dimorphic: women accumulate proportionally more in the gluteofemoral depot — buttocks, hips, thighs — than men do, a pattern set by sex hormones and reinforced across the reproductive years⁴. It is not incidental padding. Gluteofemoral fat behaves differently from abdominal fat: it is metabolically more passive day to day, and it protects by taking up and holding excess fatty acids over the long term. Higher gluteofemoral fat mass independently associates with a better lipid and glucose profile, and conditions that strip it out are associated with worse metabolic and cardiovascular risk⁵.

Which produces the awkward truth about "Ozempic butt": the depot that visibly deflates is one you would, metabolically speaking, rather keep. That is a reason to lose weight at a survivable pace, not a reason to avoid treatment.

Three things stack in that region at once — the fat compartment shrinks, the gluteal muscle underneath loses mass along with the rest of the body, and skin that was stretched around a larger volume does not retract as fast as the volume disappears. The loose-skin evidence covers that third component, and the muscle-loss guide the second.

How to avoid "Ozempic butt"

What has actually been tested is lean-mass preservation, not shape. That distinction matters, so I will keep it visible.

A 2024 review in Diabetes Care put the lean-mass figure at roughly 10%, or about 6 kg, on incretin therapy — a loss the authors compare to a decade or more of aging — and argued for resistance exercise as a deliberate adjunct rather than general advice to "stay active"⁶. Its supporting numbers: supervised resistance training programs running longer than 10 weeks produce large increases in lean mass, on the order of 3 kg, and strength gains around 25%, in both men and women. It also notes that after a low-calorie diet, combining aerobic exercise with liraglutide improved weight-loss maintenance compared with either alone⁶.

Read that carefully. It is a narrative review proposing a strategy, not a trial that randomized GLP-1 users to lifting weights and photographed the result. What it supports is that trained muscle can be added while fat is being lost. What it does not establish is that this restores gluteal contour, because contour was never measured.

The other lever is rate. Nothing in the literature isolates titration speed as a determinant of shape, but the mechanism is not mysterious: skin and soft tissue remodel slowly, and a slower descent gives them more of the thing they need, which is time. Treat that as reasoning, not evidence.

How to fix it once it has happened

Here the evidence thins to almost nothing, and the marketing does not.

• Regional fat loss cannot be reversed regionally. You cannot re-fill one depot. Gaining weight back re-fills it along with everywhere else, which is not a treatment.
• Resistance training adds muscle under the fat, not fat. Glute-focused training changes the shape of the muscle, and that is a real change — but it is building a different tissue than the one that left.
• Surgical and injectable volume restoration exists, and has not been studied in this population. There is no trial of gluteal augmentation or filler in GLP-1-treated patients, and no outcome data specific to them.
• Topical products have no evidence here at all. The one skin study circulating in this space examined facial volume in an uncontrolled 12-week design, and it is discussed on the Ozempic face page rather than this one, because it does not transfer.

Where this sits on the evidence scale

That fast weight loss deflates the gluteofemoral depot: strong, and not new. It follows from decades of adipose-tissue physiology⁵ and the sex-specific distribution of fat⁴.

That GLP-1s do this more than other weight loss does: partly. More total fat leaves, faster. The proportion lost as fat versus lean tissue looks similar to placebo in the best-controlled comparison available¹, and the systematic review found non-drug weight loss crossing the same benchmarks in nearly half of studies². Calling it a drug-specific effect overstates what the data show.

That anything reverses it: no evidence. Not creams, not supplements, not a specific exercise.

That resistance training preserves lean mass during incretin therapy: moderate, and indirect. Well supported for lean mass generally, proposed rather than proven for this drug class, and never measured as an effect on shape⁶.

What would change this page: a trial that reports regional DXA or circumference outcomes — gynoid fat, hip circumference, gluteal volume — alongside a resistance-training arm. Until someone runs it, everything written confidently about fixing "Ozempic butt" is an extrapolation from a scan of the whole body.